EXPERTS EMPLOYED BY MIUC COMMISSIONS. 165 



a year apart, with careless supervision of additions to the herd, 

 are useless in a herd that was badly infected at the beginning, 

 for tuberculosis will keep pace- with lax efforts directed against 

 it. It is not sufficient to test merely the cows that happen to 

 be in milk at the time of the test. Every dry cow should be 

 included. In an infected herd, a test once in six months is 

 regarded as necessary, followed each time by thorough disin- 

 fection of the stable. The control of tuberculosis cannot be 

 accomplished by one test carried out in a perfunctory manner, 

 but the struggle must extend over years. 



Additions to the herd must be tested with tuberculin, but 

 there is always danger that an animal, though not reacting, 

 may introduce the disease. On this account it is far better to 

 subject each animal added to the herd to a three months' quar- 

 antine with a tuberculin test at the beginning and end of this 

 period. During the period the milk may be used. 



The details concerning the tuberculin test and management 

 of the disease are given in Chap. IV. 



Bacteriological examinations. The numerical determination 

 of the bacteria in milk is made every week and constitutes a use- 

 ful check upon the methods employed in producing and hand- 

 ling the milk. A low count indicates that cleanliness has been 

 observed in the care of the utensils and in milking, and that 

 the product has been properly refrigerated. The bacterial 

 count is made of the milk when about the age that it ordinarily 

 is when it reaches the consumer. Care should be taken that 

 the sample be kept properly refrigerated until the examination 

 is made. The samples should be taken from the dairyman 

 unexpectedly as regards time and place. Thus it would be 

 impossible for him to submit extra good samples with fraud- 

 ulent intent. The numerical determination of the bacteria in 

 milk is a good safeguard against the fraudulent sale of ordi- 

 nary dirty milk under the guise of certified milk. There is a 

 very wide difference in the results that are obtained by exam- 

 ining the samples of these two classes of milk and thus fraud 

 may be detected. 



The grade of milk designated "certified" should never 

 contain more than 10,000 bacteria percc. If there is a second 



